Throw away your short belts - direct drive.

[aviphysics 9]

Love Snowygrouch for getting this idea going, but I am finding that I am slowly melting through his motor mounts. Every 8 hours of printing I have to tighten the screws holding the motors in place or it starts to fall apart. Planning to switch to Nick's design before something catastrophically fails.

 

[snowygrouch 1]

AVI,

It helps when you have a problem like that to post photographs - otherwise nobody can really tell what the problem is.

However I have never heard of anyone melting the motor mounts, and I have never even once re-tightened

mine since I put them on the machine.

So I think you have another problem here. Either you used a very low temperature variation of PLA (like PLA45

from Orbitech which gets soft enough to bend in hot tap water) or you have not printed them

with thick walls, or your steppers are overheating - possibly because of very high friction from misaligned

rods or using grease on the XY rods. Or maybe a combination of all these factors.

Without pics its nearly impossible to help.

 

[zumfab 3]

Why are you guys all cooling the motors with heatsinks? They do not seem to need extra cooling in the original setup and are even better cooled outside of the Ultimaker.

 

[anthonyt 0]

Why are you guys all cooling the motors with heatsinks? They do not seem to need extra cooling in the original setup and are even better cooled outside of the Ultimaker.

 

My motor got warmer when directly linked to the rod instead of through the short belt. I suspect there is more twisting on the motor shaft which is causing the extra heat. I use a fan to cool the motor now.

 

[Dim3nsioneer 557]

The motors shouldn't get warmer as long as you don't increase the current on the stepper drivers.

If you have to increase the motor current after a modification, you should check the mechanical resistance of the print head when pushed around while the motors are switched off. If you feel different mechanical resistance at different positions, you should check for excentricity of the pulleys and un-straightness of the rods.

 

[zumfab 3]

My main gripe with this upgrade is the extra space it takes and the vulnerability of the motors of the outside of the frame. Going right back to the start the options are limited however. Using gears or cogs would mean other problems, just replacing one issue with another. Placing the motor somewhere between the ends of the axis rod is also not possible. That pretty much leaves putting the motor on the end of them, which is exactly what people are already doing.

It does seem however that about half of the extra space needed is used up by the coupling. Not really a part that you explicitly need or want - it just happens to be used to fix two parts you do need together. If it were possible to come up with a solution that fixes the motor right to the frame that would not introduce additional slop or backlash, possibly with rubber dampening in between the motor and frame to lessen noise, I think the biggest problem would be fixed. Add an appropriately designed bracket to shield the motor from any bumps and you have all the benefits and barely any drawbacks.

 

[snowygrouch 1]

The flexible coupling is alot more than just fixing two parts together, just look at your Z stage

screw (which has a fixed coupler). They generally wobble around like crazy. Without precision machined

parts you should never direct couple motors and shafts solid (of course since its a bad idea, lots of

people do it anyway....but its a great way to put stress into the motor bearings and induce

non-linearity (because the motor will be trying to wobble about like mad).

The basic consideration here is:

1) Get rid of the short belts for total cost of about €20

2) Get rid of the short belts for a total cost of about € MORE (order pancake steppers with integrated

8mm shafts from the factory).

I chose #1....

You can do anything you want but to improve on it - will cost more.

Zapp Automation in the UK can do some custom stepper stuff, they might do you a stepper

with a shaft going right through. Since the shaft goes through the motor main bearings

its already reasonably centered. That would eliminate the coupler, but I would expect noise to

increase very noticably. (this is a Haydon Clerk pancake stepper)

You can then put dampers on the back of the steppers....but they`re about €50 each

http://www.phytron.eu/antrieb/index.php?Set_ID=165&part_ID=13

 

[zumfab 3]

The flexible coupling is alot more than just fixing two parts together, just look at your Z stage

screw (which has a fixed coupler). They generally wobble around like crazy. Without precision machined

parts you should never direct couple motors and shafts solid (of course since its a bad idea, lots of

people do it anyway....but its a great way to put stress into the motor bearings and induce

non-linearity (because the motor will be trying to wobble about like mad).

I get that However, this does not detract from the fact that you are just trying to fix two parts together. The flex joint is not needed as an integral part of the design like the motor or the axis, it is just there to compensate for the lack of accuracy when aligning the parts.

If you could either eliminate the potential misaligment in a cheap way or mitigate the problems caused by any wobble through a motor mount you could remove the coupling.

I do like the pancake steppers, but their holding torque seems not up to par, unsurprisingly. Maybe not the most relevant for the gantry steppers, but less than ideal when looking at performance upgrades.

 

[jonnybischof 60]

Don't the flexible couplings also "dampen" the motor transmission to the shaft?

If you fix the motor directly to the shaft, then the motor will take the full moment of inertia (generated by the x-y-gantry when rapidly changing directions).

I'd say that's also what increases the noise, and I'm pretty sure the motors don't like it much, either.

However, if there is a flexible coupling between these parts, it will take some of the moment of inertia away.

The cheap chinese couplings even take quite a lot - you can easily twist them when you put screwdrivers through the setscrew-holes...

I don't have my direct drive setup up and running yet, so I can't speak based on actual experience. But I have lots of different parts laying around and I have made some experiments with them.

 

[zumfab 3]

That's funny, I would say the opposite - the reduced resistance on the drive train should reduce strain in the motors. Of course, the rapid change of direction is also less dampened, but I feel those effects should be minor compared to what the motors need to fight in a regular setup.

Any flex in the couplings also seems a bad thing, as play also means loss of accuracy. Maybe removing all that flex proves to be too much as forces are more concentrated, but remember that the gantry is never directly attached - there are always rubber belts dampening the movement ever so slightly.

 

[meduza 191]

We are working on our variant of Nick's motor corner, since we want to place the motors in the back left corner instead of the back right, also made it more solid and adjusted a few things along the way.

The printer in question, a heavily modified Ultimaker was bought on this forum about a year ago, and is nowdays dubbed "The Ultimonster" at Stockholm Makerspace

 

[antiklesys 3]

Meduza, that looks an interesting remix of that design. Are the files available anywhere?

Does that mod also re-uses the standard axis?

 

[zoev89 73]

Interresting, I just mirrored Nick's design to mount it on the same location. Didn't want to move the extruder, keep it open for dual extrusion if I ever want that.

I only found out later that I will have problems with the mounting of the limit switch (the screws will hit the bracket) that is located in that corner. Don't know yet how to fix it since I am waiting for parts to arrive before I proceed.

Meduza did you buy new axis or did you use some of extension mounts?

 

[jonnybischof 60]

You could just leave the positive endstops away. You only need the negative endstops (home position) in order to home the printhead. From there, it can work with the software endstops.

 

[meduza 191]

Well, to do this mod we did not need to buy any new axis, tough that does not say much since this printer was a direct-drive since before we bought it (but had problems with not so great motor mounts and motor shaft that had taken a beating, so yes the axis are longer than original, sticks out about a centimeter into a flex coupling (we also used 25mm flex couplings, that would not fit in Nicks original design, since we had those)

The files are not yet available anywhere, still some tweaks left to do

 

[mbartels 0]

Ultimaker Original now direct driven...

Didn't want to buy new rods so i printed these spacers (http://www.thingiverse.com/thing:189742). They work perfect even without the aluminium parts. I don't think i could get it to work with the al-plates because i had to melt the front one with a lighter to make room for the pully.

As for putting the motors on the printer, i used Nicks design and it works great so far. It's a little flexible but that could be because of the 20% infill. Will maybe print a 100% infill version. We'll see...

The motors are cooler than before. Don't know why to be honest, maybe the whole printer is more square than before because i had to realign everything and now there's less resistance.... Bought heatsinks just in case...

As for inverting the x-motor: inverted cable positions instead of firmware update.

 

[zoev89 73]

Was the rod damaged at the original location where the grub screw mounted the pully? Since that 'scar' moves it might damage the slider block. Or is the damage so light that it is no issue.

I am also preparing the same modification. I don't think that the flexability in nick his design is an issue as long as the motor does not tilt. Don't think that 100% infill helps. What kind of plastic did you use?

[mbartels 0]

Was the rod damaged at the original location where the grub screw mounted the pully? Since that 'scar' moves it might damage the slider block. Or is the damage so light that it is no issue.

I am also preparing the same modification. I don't think that the flexability in nick his design is an issue as long as the motor does not tilt. Don't think that 100% infill helps. What kind of plastic did you use?

 

Well i didn't check the rod for marks and felt nothing when I wiped them clean of dust prior to reassembling the axes. So, no I don't think so.

I used PLA and get the feeling that the motors in fact do tilt a bit. The lines on printing full material don't connect completely. Just two and then there's a small gap. Then two connecting and a gap again. Got no image, sorry... But somewhere there's a 'visual guide' with print errors (haven't got the link on my iPad from which I'm writing this because it's sunny outside and I needed some fresh air and that points out as solution to tighten the short belts, which I'm not capable of anymore... :/ while printing you can watch the whole mount bending/twisting slightly. Just don't know if that's enough to interfere with the print

Right now I'm thinking about to reverse the process. Printing a 100% filled version may be worth a try though...

Well, not too happy right now. Thought it would be easier.

[zoev89 73]

Hi mbartels

That does indeed not sound promising. I think I printed my version with 30% but with ABS in this case ABS is usually more flexible.... It is good to know about the rods not being damaged though. The artefact you talk about does ring a bell and could very well be related to slack in the drive system. Which in your case would be the mount. Are the motors tight enough mounted? Interestingly nobody mentioned having this kind of problem yet.

 

[chopmeister 0]

I was afraid that Nick's design was a bit flexible so I opted to make my own https://www.youmagine.com/designs/motor-corner-alternative instead.

It's a PITA to bolt the motors, but you only have to do it once. Works for me, printed with 25% infill and no flex in it whatsoever.

 

[mbartels 0]

Hi mbartels

That does indeed not sound promising. I think I printed my version with 30% but with ABS in this case ABS is usually more flexible.... It is good to know about the rods not being damaged though. The artefact you talk about does ring a bell and could very well be related to slack in the drive system. Which in your case would be the mount. Are the motors tight enough mounted? Interestingly nobody mentioned having this kind of problem yet.

 

With the Ultimaker came extra screws for the pulleys. They got a flat head pressing against the rod. The Original ones got a little whole in the middle. I bought my UM just recently und used those of course. Don't know if they came with yours too.

Motors are mounted pretty tight, yeah.

 

I was afraid that Nick's design was a bit flexible so I opted to make my own motor corner instead.

It's a PITA to bolt the motors, but you only have to do it once. Works for me, printed with 25% infill and no flex in it whatsoever.

 

...and i know what to print while i'm at work tomorrow! thank you for your effort. Will try that... even if i have to move all the stuff to another corner

edit: i do need to flip the y-motor-cables on your design too, right?!?! the x-motor-cables already are flipped.

 

[chopmeister 0]

...and i know what to print while i'm at work tomorrow! thank you for your effort. Will try that... even if i have to move all the stuff to another corner

 

For me the back/left seems a more logical place, I doesn't look as nice having it in front, and on the back/right there's the extruder so that seemed the way to go.

 

[zoev89 73]

Back/left that is exactly my thinking as well. That is why I mirrored nicks design. I do found out that the screws of the limitswitch is in the way. Is that also the case for you?

 

With the Ultimaker came extra screws for the pulleys. They got a flat head pressing against the rod. The Original ones got a little whole in the middle. I bought my UM just recently und used those of course. Don't know if they came with yours too.

Motors are mounted pretty tight, yeah.

 

ahhhh now I know what those screws are for.... Well the machine is already running for 3 months.... Will see how that ends up.

 

[chopmeister 0]

Back/left that is exactly my thinking as well. That is why I mirrored nicks design. I do found out that the screws of the limitswitch is in the way. Is that also the case for you?

 

I removed the max limit switches ages ago.

I constantly experiment with printhead designs and therefore my print volume, so the max switches would have to also be moved all the time - much easier to just do it in firmware.

 

[mbartels 0]

Took about an hour to set everything up. Forgot about the Limitswitch screws :/ Now there's a little gap.

I'd say it won't be a problem. Only problem knowing it's there Unscrew everything again if i'm bored... it's really a PITA on some corners... but i appreciate your work! No bending, flexing or whatever. Solid as a brick.

First print running right now and it looks pretty good.